1(3), May 1984, page 3

[conclusion of Winterbauer article at top of this page]

An Authoring System for Use by Teachers of Composition

Craig Hansen and Lance Wilcox
University of Minnesota

One of the major roles for those of us attempting to develop the potentials of computers for use in composition teaching and practice is that of go-between. Between the green console and whirring disk drives of a microcomputer and your average humanities-educated writing teacher the gap yawns perilously wide. Part of this is simply a problem of language. The computer "speaks" BASIC or Fortran or Pascal; the teacher speaks English.

The problem admits of only three solutions. We, the go-betweens, can assume more or less permanently the role of translators, rendering commands from English into Computer and responses from Computer in English. Or, we can become language teachers, instructing all our composition instructors in the arcana of computer usage. Or, finally, we can teach our machines to speak English. Now to do this perfectly would require language-decoding programs of a complexity at present hardly imaginable. But we don't need this degree of sophistication just yet. For now, it will be a great boon if we can render our programs designed for use in composition instruction several degrees more user-friendly and self-instructional than they are now.

This, at any rate, is the goal of a three-year research project here at the University of Minnesota. 1 One of the original goals of the project was to evaluate commercially available microcomputer software that might aid in composition instruction. It has become almost a commonplace that commercial word-processing programs are little more than fancy typewriters, designed for secretarial work, not composing. There is, however, a program package being developed by Ruth Von Blum and Michael Cohen at UCLA, which we have been testing for the past several months. The system, WANDAH, consists of three parts: prewriting aids, revision aids, and a

COMPUTERS and COMPOSITION 1(3), May 1984, page 4

powerful word processor. This is a sophisticated, useful package for inexperienced (primarily freshman) writers.

Students in more specialized upper-division courses, however, face more complex writing challenges; and a system as broad as WANDAH, though useful, does not meet all their needs. For such courses, we need programs that go beyond word processing itself to teach specific writing sub-skills, relevant to different content areas and rhetorical problems.

Creating such programs, however, entails finessing the "translation" problem discussed at the beginning of this article. Of the three methods described there, only two are at all practicable, since writing teachers themselves can hardly be expected to learn the programming required to create such exercises. This leaves us the other two options. Either we can translate the teachers' exercises into workable programs ourselves, or we can teach our machines to do the translations for us.

At present, we at Minnesota are forced to rely on the former. Teachers using computers in their sections are detailing for us the sorts of exercises they would like to see available, and members of our research group are setting these up as programs for our microcomputers. (Reports on the design, usage, and success of these programs will be forthcoming.) Though certainly a step in the right direction, this procedure is still slow, clumsy, and inflexible. Hence the need for an authoring system of the sort we presently have under development.2 This authoring system, when completed, will allow teachers with little or no previous computer experience to "program" their lessons themselves.

Commercial authoring systems are typically inflexible programs that create true/false and multiple choice questions only. For writing instruction, these systems are of limited value.

The system we are developing should prove much more adaptable. At the most basic level, the system offers multiple choice and true or false questions which are interactive--that is, the computer responds in certain predetermined ways to student input. These kinds of structures can be useful, but are far from adequate for teaching most of the skills our composition students need to learn. The system also offers non-interactive programs that ask students to produce text in response to open-ended questions: a necessity for most prewriting lessons. Additionally, our system will offer read-only texts for instructions or samples, structures to create revision and organization guides, and "checkers" that search for variables determined by the individual instructor. All these are joined by a simple editor/word processor and accessed by simple English menus. The real key here is that none of the lessons is fixed (although we will offer "pre-fabricated" lessons to those instructors who want them). The instructor should be able to assemble a lesson using any of these capabilities in any order. The authoring system provides a range of structures; the instructor provides the content.

Thus, we hope to meet the needs of our students and instructors in our upper-division courses with software that goes beyond word processing alone. The strength of our approach, as we see it, is that the authoring system, using simple English commands, will enable more instructors to use the resources and capabilities of microcomputers to aid in teaching specific writing skills.

1. This research is funded primarily by a grant from the Fund for the Improvement of Post-Secondary Education, Lillian Bridwell and Donald Ross, co-principal investigators.

2. In creating this authoring system, we have worked closely with Samuel Sharp of the Department of Civil and Mineral Engineering.